The problem really is with the headline, not with the study or even necessarily the rest of the article. The prevalence effect that the article mentioned did appear to manifest during the study. The thing is, this was true of both motorcycles and buses. Just as recognition of motorcycles became better when the motorcycles were more frequent, the recognition of buses became better when buses were more frequent. It would have be just as (in)appropriate to title the article "'Element Of Surprise' Makes Buses Greater Traffic Hazard Than Cars: The Low-Prevalence Effect In Action". It's just poor journalism.
The better headline would have been ""Element of Surprise' Makes Drivers a Greater Hazard to Motorcyclists" since they're the ones not seeing us.